Re: [v6ops] reclassify 464XLAT as standard instead of info - double stack coexistence

[Alexandre Petrescu <[e-mail]>]

Le 25/09/2017 à 10:28, Mikael Abrahamsson a écrit :
> On Mon, 25 Sep 2017, Alexandre Petrescu wrote:
> 
>> It may also be that we do speculate too much on IPv6-only cellular 
>> networks.  We speculate these are there now, whereas I may speculate 
>> it may not really be the case.  There are some networks that are 
>> IPv6-only-to-end-user yet use GTP-IPv4.
> 
> Are you saying that people from for instance T-Mobile USA saying they're 
> doing IPv6 only GTP

No, I would like to ask T-Mobile USA whether they use GTP-IPv6?  Or 
GTP-IPv4?

[...]

> What do you mean by GTP-IPV4?

I mean GTPU protocol that uses IPv4 addresses, and carries IPv6 packets 
inside; an example packet capture of what I call GTP-IPv4 is the following:
> INBOUND>>>>>  15:47:48:883 Eventid:142004(3)
> GTPU Rx PDU, from [IPv4-address]:2152 to [IPv4-address-2]:2152 (129)
> TEID: 0x80072183, Message type: GTP_TPDU_MSG (0xFF)
> Sequence Number:: NA
> GTP HEADER FOLLOWS:
> [snip]
> GTP HEADER ENDS.
>            Payload protocol: IPv6
> PROTOCOL PAYLOAD FOLLOWS:
> fe80::b9d9:c10a:2c2d:3c3b.546 > ff02::2.546:  [udp sum ok]
> IPv6 Header Follows:
> [snip]
> PROTOCOL PAYLOAD ENDS.


> We're talking about what access service is 
> provided to the customers, and there are several that do IPv6 only GTP. 
> Yes, the GTP packets might be carried over IPv4, but the SERVICE 
> provided to the customer is IPv6 only.

Excuse me, Mikael, but it is little possible to claim "IPv6-only" when 
there is IPv4.  One can call it "IPv6-only SERVICE", but there's still 
IPv4 - so it's an "IPv4-based IPv6-only SERVICE".

>> Actually, from where I look, I see many 'double stack' networks like 
>> these.
> 
> There are plenty of examples of IPv4 only, IPv6 only, and IPv4v6 mobile 
> access networks. It's all being done. What the provider chooses to do 
> often has to do with their equipment vendor supports and how many IPv4 
> addresses they might have. Also operational practices and preferences.

Yes, the operational practices and preferences are very important.

Their consequences too.

Alex